Development of scalable coding on encrypted images using BTC for different non-overlapping block size

In this paper, the novelty of scalable coding of encrypted images (S.C.E.I) using Block Truncation Code (BTC) for different non-overlapping block sizes is investigated. Initially, the original content of the image of size 512 × 512 is split into different non-overlapping block (N.O.B.) sizes of 2 × 2, 4 × 4, 8 × 8 and 16 × 16. Then each N.O.B. size will undergo image compression using a BTC followed by encryption using pseudo-random number (PN) at the transmitter side. The original content is reconstructed at the receiver by the following stages: the transmitted content is decrypted by PN, which is shared by the transmitter, then the BTC technique for image reconstruction is applied, further there is the application of scaling and the bilinear interpolation techniques. While making an analogy, it is observed that the non-overlapping block of size 4 × 4 is best suitable for image encryption procedures/applications than the 2 × 2 size in terms of lesser memory size and storage space requirements; even though better MSE, wMSE, PSNR, wPSNR, compression ratio and bit rates are delivered by the non-overlapping block of 2 × 2 size.